CN102104447A

CN102104447A - Method for optimizing passing rate performance in wireless ad hoc network

Info

Publication number: CN102104447A
Application number: CN2011100580681A
Authority: CN
Inventors: 李波; 戴瑞龙
Original assignee: Northwestern Polytechnical University
Current assignee: Northwestern Polytechnical University
Priority date: 2011-03-10
Filing date: 2011-03-10
Publication date: 2011-06-22
Anticipated expiration: 2031-03-10
Also published as: CN102104447B

Abstract

The invention discloses a method for optimizing the throughput rate performance in a wireless self-organizing network, which overcomes the shortcomings of the traditional IEEE802.11DCF mechanism that adopts the binary exponential backoff BEB algorithm in the minimum competition window, and the processing node competition is not timely enough and too passive. Dynamically detect the number of sending nodes in the contention channel, and then calculate the next minimum contention window of the sending node after successfully completing this transmission according to the frame error rate of the RTS packet, optimize the backoff time slot, and reduce the average access of nodes channel time, the backoff mechanism is superior to the traditional binary exponential backoff BEB algorithm.

Description

A Method for Optimizing Pass Rate Performance in Wireless Ad Hoc Networks

技术领域technical field

本发明涉及通信技术领域，是一种无线自组织网络中基于最佳退避机制的变速率MAC优化通过率性能的方法。The invention relates to the technical field of communication, and relates to a method for optimizing the throughput performance of a variable-rate MAC based on an optimal backoff mechanism in a wireless self-organizing network.

背景技术Background technique

变速率MAC，即V-MAC(Variable rate Media Access Control protocol)。其算法主要思想是发送节点根据信噪比值，感知信道状态，动态改变数据分组的传输模式，以提高通过率。目前已有的变速率传输方法主要分为两类：第一类是基于接收端的变速率算法，第二类是基于发送端的变速率算法。基于接收端的变速率算法主要采用RTS/CTS交换机制，接收端决定后续的数据分组采用何种传输模式，并将此信息放在CTS中反馈给发送端。其算法主要有RBAR。在RBAR中，接收端根据接收到的RTS的信号强度来估计无线信道的质量，然后决定发送端数据分组的传输模式，并通过CTS将传输模式传给发送端；另一种基于发送端的变速率传输算法，则是由发送端通过统计先前发送数据分组的误帧率来决定增高或降低数据分组的传输速率，其算法主要有ARF，ARF是第一个实现变速率传输的MAC协议。在ARF中，若数据分组传输连续成功，发送端则提高数据分组的传输速率，否则降低数据分组的传输速率。Variable rate MAC, that is, V-MAC (Variable rate Media Access Control protocol). The main idea of the algorithm is that the sending node perceives the channel state according to the signal-to-noise ratio value, and dynamically changes the transmission mode of the data packet to improve the pass rate. The existing variable rate transmission methods are mainly divided into two categories: the first type is based on the variable rate algorithm at the receiving end, and the second type is based on the variable rate algorithm at the sending end. The variable rate algorithm based on the receiving end mainly adopts the RTS/CTS exchange mechanism. The receiving end decides which transmission mode to use for subsequent data packets, and puts this information in the CTS to feed back to the sending end. Its algorithm mainly includes RBAR. In RBAR, the receiving end estimates the quality of the wireless channel according to the received signal strength of the RTS, and then determines the transmission mode of the data packet at the sending end, and transmits the transmission mode to the sending end through CTS; another variable rate based on the sending end The transmission algorithm is to increase or decrease the transmission rate of data packets by the sender by counting the frame error rate of previously sent data packets. The algorithm mainly includes ARF, which is the first MAC protocol to realize variable rate transmission. In the ARF, if the data packets are transmitted continuously and successfully, the sender increases the transmission rate of the data packets, otherwise decreases the transmission rate of the data packets.

网络中的节点在传输分组之前，需要先经历一段随机退避时间，目的是在多个节点竞争信道的时候，保证接入的有效性。802.11DCF采用的退避算法是二进制指数退避BEB(Binary Exponential Backoff)。节点在退避前，随机产生一个退避时间BackoffTime。Nodes in the network need to experience a period of random backoff time before transmitting packets. The purpose is to ensure the validity of access when multiple nodes compete for the channel. The backoff algorithm adopted by 802.11DCF is Binary Exponential Backoff BEB (Binary Exponential Backoff). Before the node backs off, a backoff time BackoffTime is randomly generated.

BackoffTime＝Random()×aSlotTimeBackoffTime＝Random()×aSlotTime

其中，Random()是均匀分布在[0，CW-1]之间的随机整数，竞争窗口CW(Contention Window)是介于由物理层特征决定的最小竞争窗口CW_min和最大竞争窗口CW_max之间的一个整数值，即CW_min≤CW≤CW_max。aSlotTime是由物理层特性决定的一个时隙的实际长度值，对于802.11a，一个时隙的长度是9μs。退避时间是一个以时隙为单位的随机整数。Among them, Random() is a random integer uniformly distributed between [0, CW-1], and the contention window CW (Contention Window) is between the minimum contention window CW _min and the maximum contention window CW _max determined by the characteristics of the physical layer. An integer value between , that is, CW _min ≤ CW ≤ CW _max . aSlotTime is the actual length of a time slot determined by the characteristics of the physical layer. For 802.11a, the length of a time slot is 9 μs. The backoff time is a random integer in units of slots.

一个节点执行退避过程时，在每一个时隙中侦听信道的状态，如果信道闲，则将退避时间计数器减1；如果信道忙，则退避时间计数器暂停(即不再递减)，直到侦听到信道处于连续空闲状态达到DIFS时间，退避过程重新被激活，继续递减。当退避计数器递减到0时，节点就可以执行发送。当多个节点同时竞争信道时，每个节点都经过一个随机时间的退避过程，才能占据信道，这样就大大减少了冲突发生的概率。另外，通过采用退避过程中的冻结机制，使得被推迟的节点在下一轮竞争中无需再次产生一个新的随机退避时间，只需继续进行计数器递减，那么，等待时间长的节点的优先级就高于新加入的节点，就可能优先得到信道，从而维护了竞争节点之间一定的公平性。When a node performs the backoff process, it monitors the status of the channel in each time slot. If the channel is idle, the backoff time counter is decremented by 1; When the channel is in the continuous idle state and reaches the DIFS time, the backoff process is activated again and continues to decrease. When the backoff counter is decremented to 0, the node can perform the transmission. When multiple nodes compete for the channel at the same time, each node must go through a random time backoff process before occupying the channel, which greatly reduces the probability of collisions. In addition, by adopting the freezing mechanism in the backoff process, the postponed node does not need to generate a new random backoff time in the next round of competition, but only needs to continue to decrement the counter, then the priority of the node with a long waiting time is high Because of the newly added nodes, it is possible to get the channel first, thus maintaining a certain fairness among competing nodes.

在通信过程中，竞争窗口CW的初始值为CW_min，如果一个节点传输数据失败，则CW需要加倍，直至加倍到CW_max，即CW＝2^m·CW_min，其中m为重传次数。当CW的值增加到CW_max后，再次重传的竞争窗口维持CW_max不变，直到该节点发送成功，或者达到最大重传次数，CW将被重新置为CW_min，如图1所示。During the communication process, the initial value of the contention window CW is CW _min . If a node fails to transmit data, the CW needs to be doubled until it doubles to CW _max , that is, CW=2 ^m ·CW _min , where m is the number of retransmissions. When the value of CW increases to CW _max , the contention window for retransmission remains unchanged until the node sends successfully or reaches the maximum number of retransmissions, and _CW will be reset to CW _min , as shown in Figure 1.

IEEE 802.11DCF的退避机制中，采用固定的最小竞争窗口，当竞争信道的发送节点个数较少的时候，最小竞争窗口相对较大，节点接入信道的平均时间较长，对信道的利用率是种浪费；当竞争信道的发送节点个数较多的时候，最小竞争窗口相对较小，增加了RTS分组的冲突概率，降低了网络的通过率。In the backoff mechanism of IEEE 802.11DCF, a fixed minimum contention window is adopted. When the number of sending nodes competing for the channel is small, the minimum contention window is relatively large, and the average time for nodes to access the channel is long, and the utilization rate of the channel is relatively large. It is a waste; when there are many sending nodes competing for the channel, the minimum contention window is relatively small, which increases the collision probability of RTS packets and reduces the network throughput rate.

发明内容Contents of the invention

为了克服现有技术信道利用率低或者网络通过率低的不足，本发明提供一种基于最佳退避机制的变速率MAC，即OV-MAC(Variable rate Media Access Control protocol with Optimal Backoff)，对网络通过率有明显改善。In order to overcome the shortcomings of low channel utilization or low network throughput in the prior art, the present invention provides a variable rate MAC based on an optimal backoff mechanism, that is, OV-MAC (Variable rate Media Access Control protocol with Optimal Backoff), for network The pass rate has improved significantly.

本发明解决其技术问题所采用的技术方案是：基于最佳退避机制的变速率MAC，在接收节点成功接收到RTS分组后，接收节点根据RTS分组的信噪比，计算IEEE802.11a的8种速率模式的误帧率，通过公式选择数据分组的最佳传输速率，且根据网络中等待发送数据的节点个数，计算出本次成功传输后下次退避的最小竞争窗口。The technical solution adopted by the present invention to solve the technical problem is: based on the variable rate MAC of the optimal backoff mechanism, after the receiving node successfully receives the RTS packet, the receiving node calculates the 8 types of IEEE802.11a according to the signal-to-noise ratio of the RTS packet. The frame error rate of the rate mode selects the optimal transmission rate of the data packet through the formula, and calculates the minimum contention window for the next back-off after the successful transmission according to the number of nodes waiting to send data in the network.

判断当前网络中等待发送数据的节点个数原理为：对于发送节点i，当监听到发送节点j发送的数据分组时，读出节点j数据中的B_j，修改当前状态的参数k，k表示网络中等待发送数据的节点个数；(B_j＝1表示节点j还有数据分组等待发送；B_j＝0表示节点j无数据分组等待发送)。The principle of judging the number of nodes waiting to send data in the current network is: for sending node i, when listening to the data packet sent by sending node j, read B _j in the data of node j, and modify the parameter k of the current state, k means The number of nodes waiting to send data in the network; (B _j =1 indicates that node j still has data packets waiting to be sent; B _j =0 indicates that node j has no data packets waiting to be sent).

为了支持上述提出的算法，我们重新定义了CTS的帧格式，如图4所示。在CTS帧格式中分别加入1个字节的“传输模式”域和1个字节的“最小竞争窗口”域。对于“传输模式”域的8位，前5位暂不使用，后3位表示IEEE 802.11a的8种传输模式；对于“最小竞争窗口”域的8位，只使用最后1位表示B_j的值，其余的暂不使用。In order to support the algorithm proposed above, we redefine the frame format of CTS, as shown in Fig. 4. In the CTS frame format, a 1-byte "transmission mode" field and a 1-byte "minimum contention window" field are respectively added. For the 8 bits of the "transmission mode" field, the first 5 bits are not used temporarily, and the last 3 bits represent the 8 transmission modes of IEEE 802.11a; for the 8 bits of the "minimum contention window" field, only the last 1 bit is used to represent the B _j value, and the rest are not used for now.

具体步骤如下：Specific steps are as follows:

情况1：发送节点i第一次发送数据分组。Case 1: The sending node i sends a data packet for the first time.

步骤1：发送节点i在第一次发送数据分组时，采用802.11DCF二进制指数退避BEB(Binary Exponential Backoff)，在退避时间结束后，发送RTS分组；Step 1: Sending node i uses 802.11DCF Binary Exponential Backoff BEB (Binary Exponential Backoff) when sending data packets for the first time, and sends RTS packets after the backoff time ends;

步骤2：若发送节点i正确收到CTS分组，读出其携带的传输模式值和最小竞争窗口值，使用该传输模式发送数据分组，并将

值保存起来；如果节点i还有后续数据分组，将B_j＝1写入数据分组中，否则将B_j＝0写入数据分组；若发送节点没有收到CTS分组，则将竞争窗口加倍，采用BEB重选退避值，进入退避过程，返回步骤1；Step 2: If the sending node i receives the CTS packet correctly, read the transmission mode value and the minimum contention window carried by it value, use this transmission mode to send data packets, and

Save the value; if node i still has subsequent data packets, write B _j =1 in the data packets, otherwise write B _j =0 into the data packets; if the sending node does not receive the CTS packet, then the contention window will be doubled, Use BEB to reselect the backoff value, enter the backoff process, and return to step 1;

步骤3：若发送节点i正确接收到ACK确认分组，则说明本次传输成功，将最小竞争窗口修改为

值，回到退避状态，等待传输下一个数据分组。若发送节点没有正确接收到ACK，表明此次传输失败，则将竞争窗口加倍，采用BEB重选退避值，在等待信道持续空闲1个DIFS帧间间隔后，进入退避过程，返回步骤1。Step 3: If the sending node i correctly receives the ACK confirmation packet, it means that the transmission is successful, and the minimum contention window is changed to

value, return to the backoff state, and wait for the next data packet to be transmitted. If the sending node does not receive the ACK correctly, indicating that the transmission failed, the contention window will be doubled, and the BEB reselection backoff value will be used. After waiting for the channel to remain idle for 1 DIFS interframe interval, enter the backoff process and return to step 1.

情况2：发送节点i第二次发送数据分组及第n次发送数据分组，n＞2。Case 2: the sending node i sends the data packet for the second time and the data packet for the nth time, n>2.

步骤1：发送节点i在第二次发送数据分组时，采用情况1中第一次发送数据分组成功后计算出的最小竞争窗口选择退避时间，在退避时间结束后，发送RTS分组；若发送节点i在第n次发送数据分组时，采用第n-1次发送数据分组成功后计算出的最小竞争窗口选择退避时间，在退避时间结束后，发送RTS分组；Step 1: When the sending node i sends the data packet for the second time, it selects the backoff time using the minimum contention window calculated after the first successful sending of the data packet in Case 1, and sends the RTS packet after the backoff time ends; if the sending node i When sending data packets for the nth time, use the minimum contention window calculated after the successful sending of data packets for the n-1th time to select the backoff time, and send the RTS packet after the backoff time ends;

步骤2、3分别同情况1中的步骤2、3。Steps 2 and 3 are the same as Steps 2 and 3 in Case 1 respectively.

情况3：接收节点正确收到给自己的RTS分组。Case 3: The receiving node correctly receives the RTS packet addressed to itself.

步骤1：接收节点没有正确接收RTS分组，不做任何反应；正确接收RTS分组，计算传输数据分组的最佳速率；Step 1: The receiving node does not receive the RTS packet correctly, and does not make any response; correctly receives the RTS packet, and calculates the optimal rate for transmitting data packets;

假设当前时隙节点将要在时间T内发送或重发数据分组(包括传输RTS、CTS和ACK的时间在内)，设发送数据分组的时间为T_L，无论采取何种速率发送数据分组，T_L为固定长度。此时，该时隙的通过率可以用下式表示：Assuming that the node in the current time slot will send or retransmit data packets within time T (including the time of transmitting RTS, CTS and ACK), let the time of sending data packets be T _L , no matter what rate is used to send data packets, T _L is a fixed length. At this time, the pass rate of this time slot can be expressed by the following formula:

$\frac{((11 - - {p p}_{c c,, i i})) ((11 - - {P P}_{e e,, i i})) \cdot &Center Dot; {R R}_{i i} \cdot &Center Dot; {T T}_{L L}}{T T} &DoubleRightArrow; &DoubleRightArrow; {R R}_{i i} ((11 - - {P P}_{e e,, i i})) - - - - - - ((11))$

其中p_c，i表示节点i的冲突概率，当每个节点的数据分组都相同时它可以被看作常数1-e^1/K，并且

也是一个常数。R_i代表节点i的传输模式对应的传输速率，P_e，i代表在该传输模式下数据分组的误帧率。可以看出R_i(1-P_e，i)是决定系统通过率的关键因素，尽可能的使R_i(1-P_e，i)最大，将得到最好的通过率，因此，使R_i(1-P_e，i)最大所对应的R_i就表示传输数据分组的最佳速率。where p _c,i represents the collision probability of node i, which can be regarded as a constant 1-e ^1/K when the data packets of each node are the same, and

is also a constant. R _i represents the transmission rate corresponding to the transmission mode of node i, and P _e,i represents the frame error rate of data packets in this transmission mode. It can be seen that R _i (1-P _{e, i} ) is the key factor determining the pass rate of the system, making R _i (1-P _{e, i} ) as large as possible will get the best pass rate, therefore, make R The R _i corresponding to the maximum _i (1-P _{e, i} ) represents the optimal rate for transmitting data packets.

下面我们给出误帧率P_e，i的计算方法。Below we give the calculation method of the frame error rate P _e,i .

对于BPSK调制，误比特率

计算方法为：For BPSK modulation, the bit error rate

The calculation method is:

${P P}_{b b}^{((22))} = = {P P}_{22} = = Q Q ((\sqrt{22 {γ γ}_{i i}})) - - - - - - ((22))$

对于QPSK、16-QAM和64-QAM调制，高斯白噪声环境下，误比特率

计算方法(M_i表示指定调制方式发送符号的种类个数)如下：For QPSK, 16-QAM, and 64-QAM modulation, the bit error rate in Gaussian white noise environment

The calculation method (M _i represents the number of types of symbols sent by the specified modulation mode) is as follows:

${P P}_{b b}^{(({M m}_{i i}))} \approx \approx \frac{11}{{log log}_{22} {M m}_{i i}} \cdot &Center Dot; {P P}_{s the s,, i i} (({M m}_{i i},, {γ γ}_{i i})) - - - - - - ((33))$

首先计算误符号率P_s，i(M_i，γ_i)：First calculate the symbol error rate P _s,i (M _i ,γ _i ):

${P P}_{s the s,, i i} (({M m}_{i i},, {γ γ}_{i i})) = = 11 - - {[[11 - - 22 \cdot \cdot ((11 - - \frac{11}{\sqrt{{M m}_{i i}}})) \cdot \cdot Q Q ((\sqrt{\frac{33}{{M m}_{i i} - - i i} \cdot \cdot {γ γ}_{i i}}))]]}^{22} - - - - - - ((44))$

其中γ_i为每符号平均信噪比，E_b，i表示一个符号的平均能量，N_0，i表示噪声的功率谱密度。where _γi is the average signal-to-noise ratio per symbol, E _b,i represents the average energy of a symbol, N _0,i represents the power spectral density of the noise.

其中的Q函数为 $Q (x) = {&Integral;}_{x}^{\infty} \frac{1}{\sqrt{2 π}} \cdot e^{- y^{2} / 2} dy - - - (5)$ where the Q function is $Q (x) = {&Integral;}_{x}^{\infty} \frac{1}{\sqrt{2 π}} &Center Dot; e^{- {they}^{2} / 2} dy - - - (5)$

式(5)中x表示一个未知变量。In formula (5), x represents an unknown variable.

然后计算误帧率

由于卷积码的误比特之间相互关联，因此不能使用其他分组码由信道误比特率直接计算误帧率，而应该通过数学方法分析维特比译码的特性，给出卷积码误帧率的上限。Then calculate the frame error rate

Since the bit errors of convolutional codes are interrelated, other block codes cannot be used to directly calculate the frame error rate from the channel bit error rate, but the characteristics of Viterbi decoding should be analyzed mathematically to give the frame error rate of the convolutional code upper limit.

假定使用维特比硬判决译码方式，计算传输模式为m，整数m表示IEEE 802.11a标准中定义的8种传输模式所对应的序号。l₁为前导符号与PLCP头部比特总数，l₂为MPDU的比特总数。长度为l比特(l＝l₁+l₂)的数据分组的误帧率上限公式如下：Assuming that the Viterbi hard-decision decoding method is used, the calculated transmission mode is m, and the integer m represents the sequence numbers corresponding to the eight transmission modes defined in the IEEE 802.11a standard. l ₁ is the total number of bits in the preamble symbol and the PLCP header, and l ₂ is the total number of bits in the MPDU. The formula for the upper limit of the frame error rate of a data packet with a length of l bits (l=l ₁ +l ₂ ) is as follows:

${P P}_{e e,, i i}^{m m} ((l l)) \leq \leq 11 - - {[[11 - - {P P}_{u u}^{m m} (({l l}_{11}))]]}^{{l l}_{11}} {[[11 - - {P P}_{u u}^{m m} (({l l}_{22}))]]}^{{l l}_{22}} - - - - - - ((66))$

${P P}_{u u}^{m m} ((l l)) \leq \leq {Σ Σ}_{d d = = {d d}_{free free}}^{\infty \infty} {a a}_{d d} \cdot &Center Dot; {P P}_{d d} - - - - - - ((77))$

式(7)中，为首次事件错误概率的联合界。In formula (7), is the joint bound for the first-event error probability.

式(8)中，d_free是模式m下卷积码的自由距离；a_d是重量为d的错误事件的总数，可查表得到；P_d是与正确路径距离为d的错误路径被维特比译码器选择的概率，p_m为使用传输模式为m的误比特率。以上就是本方法中关于误帧率和速率选择的计算方法。In formula (8), d _free is the free distance of the convolutional code in mode m; a _d is the total number of error events with a weight of d, which can be obtained by looking up the table; P _d is the error path with a distance of d from the correct path. than the probability chosen by the decoder, p _m is the bit error rate using the transmission mode m. The above is the calculation method of frame error rate and rate selection in this method.

步骤2：根据步骤1中数据分组选择的传输模式对应的误帧率P_e，i，计算最小竞争窗口

包括以下步骤：Step 2: Calculate the minimum contention window according to the frame error rate P _e,i corresponding to the transmission mode selected by the data packet in step 1

Include the following steps:

$\overset{&OverBar; &OverBar;}{{P P}_{e e,, i i}} ((t t + + 11)) = = {β β}_{11} \cdot &Center Dot; \overset{&OverBar; &OverBar;}{{P P}_{e e,, i i}} ((t t)) + + ((11 - - {β β}_{11})) \cdot \cdot {P P}_{e e,, i i} - - - - - - ((99))$

式(9)中β₁表示一个取值在0～1之间的参数，

表示节点i在t时刻的平均误帧率，

表示节点i在t+l时刻的平均误帧率。In formula (9), β ₁ represents a parameter whose value is between 0 and 1,

Indicates the average frame error rate of node i at time t,

Indicates the average frame error rate of node i at time t+l.

${τ τ}^{* *}_{ap ap,, i i} = = \frac{11}{\sqrt{\frac{{T T}_{c c}^{* *}}{22} \cdot \cdot N N}} - - - - - - ((1010))$

式(10)中τ^* _ap，i表示节点i的近似最佳工作点，N表示等待发送数据的节点个数(N＝k)，

表示归一化的冲突持续的平均时间。In formula (10), τ ^* _{ap, i} represents the approximate optimal working point of node i, N represents the number of nodes waiting to send data (N=k),

Indicates the normalized average duration of conflicts.

${T T}_{c c}^{* *} = = \frac{{T T}_{c c}}{σ σ} - - - - - - ((1111))$

式中T_c表示信道感知到冲突持续的平均时间，T_c＝T_DIFS+T_RTS(T_DIFS表示帧间隔DIFS的时间长度，T_RTS表示传输RTS分组的时间长度)，σ表示一个空时隙。In the formula, T _c represents the average time for the channel to perceive the conflict, T _c = T _DIFS + T _RTS (T _DIFS represents the time length of the frame interval DIFS, T _RTS represents the time length of the transmission RTS packet), σ represents an empty time slot .

${p p}^{* *}_{c c,, i i} \approx \approx 11 - - \frac{{e e}^{- - 11 / / K K}}{((11 - - {τ τ}_{ap ap,, i i}^{* *}))} - - - - - - ((1212))$

式(12)中P^* _c，i表示节点i的冲突概率， $K = \sqrt{T_{c}^{*} / 2} .$ In formula (12), P ^* _{c, i} represents the collision probability of node i, $K = \sqrt{T_{c}^{*} / 2} .$

${p p}_{f f,, i i}^{* *} = = {p p}^{* *}_{c c,, i i} + + ((11 - - {p p}^{* *}_{c c,, i i})) \cdot &Center Dot; \overset{&OverBar; &OverBar;}{{P P}_{e e,, i i}} - - - - - - ((1313))$

式(13)中

表示节点i的总错误概率。In formula (13)

Indicates the total error probability of node i.

${W W}_{i i}^{* *} \approx \approx \frac{22 ((11 - - 22 {p p}_{f f,, i i}^{* *}))}{{τ τ}^{* *}_{ap ap,, i i} \cdot \cdot ((11 - - 22 {p p}_{f f,, i i}^{* *})) + + {τ τ}^{* *}_{ap ap,, i i} \cdot &Center Dot; {p p}_{f f,, i i}^{* *} \cdot &Center Dot; [[11 - - {((22 {p p}_{f f,, i i}^{* *}))}^{{m m}_{i i}}]]} - - - - - - ((1414))$

式(14)中

表示节点i的最小竞争窗口大小，m_i表示节点i的本次传输退避时隙个数。利用公式(14)可以计算出节点i成功发送本次分组后，下次竞争信道时的最小竞争窗口。In formula (14)

Indicates the minimum contention window size of node i, and m _i indicates the number of backoff slots for this transmission of node i. Formula (14) can be used to calculate the minimum contention window for the next channel competition after node i successfully sends this packet.

步骤3：将计算出的最佳传输模式与最小竞争窗口

分别写入CTS头中新引入的“传输模式”域和“最小竞争窗口”域，接收节点发送CTS分组；Step 3: Compare the calculated optimal transmission mode with the minimum contention window

Write the newly introduced "transmission mode" field and "minimum contention window" field in the CTS header respectively, and the receiving node sends the CTS packet;

步骤4：若接收节点正确接收到数据分组，就会给发送节点返回ACK确认分组。若接收节点没有正确接收到数据分组，则不做任何反应。Step 4: If the receiving node correctly receives the data packet, it will return an ACK confirmation packet to the sending node. If the receiving node does not receive the data packet correctly, no response will be made.

本发明的有益效果是：The beneficial effects of the present invention are:

仿真实验表明，基于最佳退避机制的变速率MAC在通过率性能上优于支持变速率的V-MAC。V-MAC协议中采用固定的最小竞争窗口值，当竞争信道的发送节点个数较少的时候，最小竞争窗口相对较大，节点接入信道的平均时间较长，对信道的利用率是种浪费；当竞争信道的发送节点个数较多的时候，最小竞争窗口相对较小，增加了RTS分组的冲突概率，降低了网络的通过率。而OV-MAC协议可以动态的检测竞争信道的发送节点个数，然后根据RTS分组的误帧率，计算出该发送节点在成功完成本次传输后下次的最小竞争窗口，优化退避时隙，减小节点平均接入信道的时间，从而提高网络的通过率性能。Simulation experiments show that the variable-rate MAC based on the optimal backoff mechanism is better than the V-MAC that supports variable rates in throughput performance. The V-MAC protocol adopts a fixed minimum contention window value. When the number of sending nodes competing for the channel is small, the minimum contention window is relatively large, and the average time for nodes to access the channel is long, which is an important factor for channel utilization. Waste; when the number of sending nodes competing for the channel is large, the minimum contention window is relatively small, which increases the collision probability of RTS packets and reduces the network throughput rate. The OV-MAC protocol can dynamically detect the number of sending nodes competing for the channel, and then calculate the next minimum contention window of the sending node after successfully completing this transmission according to the frame error rate of the RTS packet, and optimize the backoff time slot. Reduce the average access time of nodes to the channel, thereby improving the throughput performance of the network.

在非饱和状态下，均匀随机放置收发节点对，每对收发节点相距大约100米，随着收发节点对个数的增加，对比具有最佳退避机制的OV-MAC协议和V-MAC协议的通过率性能，如图5所示。In the unsaturated state, the transceiver node pairs are evenly placed randomly, and each pair of transceiver nodes is about 100 meters apart. With the increase of the number of transceiver node pairs, the OV-MAC protocol and the V-MAC protocol with the best backoff mechanism are compared Rate performance, as shown in Figure 5.

从图5中可以看出，OV-MAC的网络通过率明显高于V-MAC。因为V-MAC协议采用固定的窗口大小，在发送节点少的时候，最小竞争窗口相对较大，节点接入信道的平均时间相对较长，对无线信道是一种浪费，而OV-MAC可以动态改变最小竞争窗口大小，在发送节点少时，降低最小竞争窗口的值，减小了节点接入网络的平均时间，提高了网络通过率。因此可以看出OV-MAC协议中的退避机制优于传统的二进制指数退避BEB(Binary Exponential Backoff)算法。It can be seen from Figure 5 that the network pass rate of OV-MAC is significantly higher than that of V-MAC. Because the V-MAC protocol uses a fixed window size, when there are few sending nodes, the minimum contention window is relatively large, and the average time for nodes to access the channel is relatively long, which is a waste of wireless channels, while OV-MAC can dynamically Change the size of the minimum contention window, reduce the value of the minimum contention window when there are few sending nodes, reduce the average time for nodes to access the network, and improve the network throughput rate. Therefore, it can be seen that the backoff mechanism in the OV-MAC protocol is superior to the traditional binary exponential backoff BEB (Binary Exponential Backoff) algorithm.

下面结合附图和实施例对本发明进一步说明。The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

附图说明Description of drawings

图1是竞争窗口示例示意图；Figure 1 is a schematic diagram of an example of a competition window;

图2是OV-MAC实现流程图；Figure 2 is a flow chart of OV-MAC implementation;

图3是PPDU帧格式示意图；Fig. 3 is a schematic diagram of a PPDU frame format;

图4是基于最佳退避机制的CTS分组结构示意图；FIG. 4 is a schematic diagram of a CTS grouping structure based on an optimal backoff mechanism;

图5是OV-MAC和V-MAC的性能对比示意图。Fig. 5 is a schematic diagram of performance comparison between OV-MAC and V-MAC.

具体实施方式Detailed ways

基于最佳退避机制的变速率MAC，在接收节点成功接收到RTS分组后，接收节点根据RTS分组的信噪比，计算IEEE 802.11a的8种速率模式的误帧率，通过公式选择数据分组的最佳传输速率，且根据网络中等待发送数据的节点个数，计算出本次成功传输后下次退避的最小竞争窗口。具体流程图如图2所示：Based on the variable rate MAC of the optimal backoff mechanism, after the receiving node successfully receives the RTS packet, the receiving node calculates the frame error rate of the eight rate modes of IEEE 802.11a according to the signal-to-noise ratio of the RTS packet, and selects the frame error rate of the data packet through the formula The optimal transmission rate, and according to the number of nodes waiting to send data in the network, calculate the minimum contention window for the next backoff after this successful transmission. The specific flow chart is shown in Figure 2:

为了便于性能对比与分析，结果中只采用18Mbps、36Mbps和54Mbps三种速率传输数据分组，RTS、CTS以及ACK分组采用最低速率传输，即6Mbps。仿真场景采用均匀随机的收发节点对。表1为设置的仿真参数。In order to facilitate performance comparison and analysis, only three data rates of 18Mbps, 36Mbps and 54Mbps are used in the results to transmit data packets, and RTS, CTS and ACK packets are transmitted at the lowest rate, that is, 6Mbps. The simulation scenario uses a uniform random pair of sending and receiving nodes. Table 1 is the simulation parameters set.

表1系统仿真参数Table 1 System Simulation Parameters

具体步骤如下：Specific steps are as follows:

情况1：发送节点1第一次发送数据分组。Case 1: Sending node 1 sends a data packet for the first time.

步骤1：发送节点1在帧间间隔DIFS结束后，采用BEB算法随机产生退避时间(随机选择10个时隙)，在退避时间结束后，发送RTS分组；Step 1: After the inter-frame interval DIFS ends, the sending node 1 uses the BEB algorithm to randomly generate a backoff time (randomly selects 10 time slots), and sends an RTS packet after the backoff time ends;

步骤2：发送节点1正确收到CTS分组，读出其携带的传输模式值(36Mbps)和最小竞争窗口

使用该传输模式发送数据分组，并将

值保存起来，检测到本节点队列中有数据分组等待发送，将B_j＝1写入数据分组中，发送数据分组；Step 2: Sending node 1 correctly receives the CTS packet, and reads out the transmission mode value (36Mbps) and the minimum contention window carried by it

Send data packets using this transfer mode, and

Save the value, detect that there is a data packet waiting to be sent in the queue of this node, write B _j = 1 into the data packet, and send the data packet;

步骤3：发送节点1正确接收到ACK确认分组，将最小竞争窗口修改为14，回到退避状态，等待传输下一个数据分组。Step 3: The sending node 1 correctly receives the ACK confirmation packet, modifies the minimum contention window to 14, returns to the backoff state, and waits for the transmission of the next data packet.

情况2：发送节点1第二次发送数据分组及第n次发送数据分组(n＞2)。Case 2: sending node 1 sends the data packet for the second time and sends the data packet for the nth time (n>2).

步骤1：发送节点1在帧间间隔DIFS结束后，采用情况1中第一次发送数据分组成功后计算出的最小竞争窗口选择退避时间(随机选择3个时隙)，在退避时间结束后，发送RTS分组；Step 1: After the end of the inter-frame interval DIFS, the sending node 1 adopts the minimum contention window calculated after the first successful sending of data packets in case 1 Select the backoff time (select 3 time slots at random), and send the RTS packet after the backoff time ends;

步骤2：发送节点1正确收到CTS分组，读出其携带的传输模式值(54Mbps)和最小竞争窗口

使用该传输模式发送数据分组，并将值保存起来，检测到本节点队列中无数据分组等待发送，将B_j＝0写入数据分组中，发送数据分组；Step 2: Sending node 1 correctly receives the CTS packet, reads out the transmission mode value (54Mbps) and the minimum contention window carried by it

Send data packets using this transfer mode, and Save the value, detect that there is no data packet waiting to be sent in the queue of this node, write B _j = 0 into the data packet, and send the data packet;

步骤3：发送节点1正确接收到ACK确认分组，将最小竞争窗口修改为16，回到退避状态，等待传输下一个数据分组。Step 3: The sending node 1 correctly receives the ACK confirmation packet, modifies the minimum contention window to 16, returns to the backoff state, and waits for the transmission of the next data packet.

情况3：接收节点2正确收到给自己的RTS分组。Case 3: The receiving node 2 correctly receives the RTS packet addressed to itself.

步骤1：接收节点2正确接收RTS分组，计算出RTS的信噪比SNR＝18.6058dB；根据信噪比计算仿真中要求的3种传输模式的误帧率P_e，i，从而计算传输数据分组的最佳速率；Step 1: The receiving node 2 receives the RTS packet correctly, and calculates the signal-to-noise ratio of RTS SNR=18.6058dB; calculates the frame error rate P _e,i of the three transmission modes required in the simulation according to the signal-to-noise ratio, and thus calculates the transmission data packet the best rate of

首先计算54Mbps模式下的误帧率；First calculate the frame error rate in 54Mbps mode;

根据公式(3)得到误符号率：P_s，i(M_i，γ_i)＝0.107355；According to the formula (3), the symbol error rate is obtained: P _s,i (M _i ,γ _i )=0.107355;

根据公式(2)得到误比特率：P_b＝0.0178926；According to the formula (2), the bit error rate is obtained: P _b =0.0178926;

根据公式(8)得到l₁和l₂的P_d值：Obtain the P _d value of l ₁ and l ₂ according to formula (8):

${P P}_{d d} (({l l}_{11})) = = \{\begin{matrix} 00 & {a a}_{d d} = = 1111 \\ 00 & {a a}_{d d} = = 00 \\ 00 & {a a}_{d d} = = 3838 \\ 00 & {a a}_{d d} = = 00 \end{matrix}$

P_d(l₂)＝{5.57554e-05a_d＝8P _d (l ₂ )={5.57554e-05a _d =8

根据公式(7)得到： $P_{u}^{m} (l_{1}) = 0;$ $P_{u}^{m} (l_{2}) = 0.000446044;$ According to formula (7): $P_{u}^{m} (l_{1}) = 0;$ $P_{u}^{m} (l_{2}) = 0.000446044;$

根据公式(6)可得误帧率P_e，i：P_e，i＝0.53742；According to the formula (6), the frame error rate P _e,i can be obtained: P _e,i =0.53742;

同理可以得到36Mbps和18Mbps模式下的误帧率P_e，i，即：Similarly, the frame error rate P _e,i in the 36Mbps and 18Mbps modes can be obtained, namely:

${P P}_{e e,, i i} = = \{\begin{matrix} 0.53742 0.53742 & {R R}_{i i} = = 5454 Mbps Mbps \\ 1.3207 1.3207 e e - - 0808 & {R R}_{i i} = = 3636 Mbps Mbps \\ 00 & {R R}_{i i} = = 1818 Mbps Mbps \end{matrix}$

根据公式(1)可以得到：According to formula (1), we can get:

${R R}_{i i} \cdot \cdot ((11 - - {P P}_{e e,, i i})) = = \{\begin{matrix} 24.97932 24.97932 & {R R}_{i i} = = 5454 Mbps Mbps \\ 35.9999995 35.9999995 & {R R}_{i i} = = 3636 Mbps Mbps \\ 1818 & {R R}_{i i} = = 1818 Mbps Mbps \end{matrix}$

因此本次传输的最佳速率为36Mbps。Therefore, the optimal rate for this transmission is 36Mbps.

步骤2：根据选择的36Mbps传输模式对应的误帧率P_e，i＝1.3207e-08，计算最小竞争窗口 Step 2: Calculate the minimum contention window according to the frame error rate P _e,i = 1.3207e-08 corresponding to the selected 36Mbps transmission mode

根据公式(9)得到： $\overset{&OverBar;}{P_{e, i}} (t + 1) = 1.3207 e - 08;$ According to the formula (9): $\overset{&OverBar;}{P_{e, i}} (t + 1) = 1.3207 e - 08;$

T_c＝T_DIFS+T_RTS＝86us，σ＝9us，因此 $T_{c}^{*} = 9.5556;$ T _c ＝T _DIFS +T _RTS ＝86us, σ＝9us, so $T_{c}^{*} = 9.5556;$

N＝k＝6，因此 ${τ^{*}}_{ap, i} = \frac{1}{\sqrt{\frac{T_{c}^{*}}{2}} \cdot N} = 0.0762491;$ N=k=6, so ${τ^{*}}_{ap, i} = \frac{1}{\sqrt{\frac{T_{c}^{*}}{2}} &Center Dot; N} = 0.0762491;$

$K = \sqrt{\frac{T_{c}^{*}}{2}} = 10.929,$ 因此p^* _c，i≈0.314894； $K = \sqrt{\frac{T_{c}^{*}}{2}} = 10.929,$ Therefore p ^* _c,i ≈ 0.314894;

$p_{f, i}^{*} = 0.314894009048195;$ mi＝10；因此 $W_{i}^{*} \approx 14 .$ $p_{f, i}^{*} = 0.314894009048195;$ mi = 10; therefore $W_{i}^{*} \approx 14 .$

步骤3：将计算出的最佳传输模式与最小竞争窗口分别写入CTS头中新引入的“传输模式”域和“最小竞争窗口”域。接收节点发送CTS分组；Step 3: Compare the calculated optimal transmission mode with the minimum contention window Write the newly introduced "Transfer Mode" field and "Minimum Contention Window" field in the CTS header respectively. The receiving node sends a CTS packet;

步骤4：接收节点2正确接收到数据分组，给发送节点返回ACK确认分组。Step 4: The receiving node 2 correctly receives the data packet, and returns an ACK confirmation packet to the sending node.

Claims

1. a method for optimizing throughput performance in a wireless ad hoc network, characterized in that it may further comprise the steps:

Case 1. Sending node i sends a data packet for the first time:

Step 1: When the sending node i sends the data packet for the first time, it uses the binary exponential backoff BEB, and sends the RTS packet after the backoff time ends;

Step 2: If the sending node i receives the CTS packet correctly, read the transmission mode value and the minimum contention window carried by it

value, use this transmission mode to send data packets, and

Step 3: If the sending node i correctly receives the ACK confirmation packet, it means that the transmission is successful, and the minimum contention window is changed to

value, return to the backoff state, and wait for the next data packet to be transmitted; if the sending node does not receive the ACK correctly, indicating that the transmission failed, the contention window will be doubled, and the BEB reselection backoff value will be used, and the waiting channel will continue to be idle for 1 DIFS After the inter-frame interval, enter the backoff process and return to step 1.

Case 2, sending node i sends data packets for the second time and sends data packets for the nth time, n>2:

Step 1: When the sending node i sends the data packet for the second time, it selects the backoff time using the minimum contention window calculated after the first successful sending of the data packet in Case 1, and sends the RTS packet after the backoff time ends; if the sending node When i send the data packet for the nth time, use the minimum contention window calculated after the n-1th data packet is successfully sent to select the backoff time, and send the RTS packet after the backoff time ends;

Steps 2 and 3 are the same as steps 2 and 3 in case 1 respectively;

Case 3. The receiving node correctly receives the RTS packet addressed to itself:

Step 1: The receiving node does not receive the RTS packet correctly, and does not make any response; correctly receives the RTS packet, and calculates the optimal rate for transmitting data packets;

Assuming that the node in the current time slot will send or retransmit data packets within time T, let the time for sending data packets be T _L , the pass rate of this time slot can be expressed by the following formula:

\frac{((11 - - {p p}_{c c,, i i})) ((11 - - {P P}_{e e,, i i})) \cdot \cdot {R R}_{i i} \cdot \cdot {T T}_{L L}}{T T} &DoubleRightArrow; &DoubleRightArrow; {R R}_{i i} ((11 - - {P P}_{e e,, i i})) - - - - - - ((11))

Among them, p _{c, i} represent the collision probability of node i, R _i represents the transmission rate corresponding to the transmission mode of node i, P _{e, i} represent the frame error rate of data packets in this transmission mode, so that R _i (1-P _{e , i} ) the maximum corresponding R _i just represents the best rate for transmitting data packets;

Below we give the calculation method of the frame error rate P _e,i :

For BPSK modulation, the bit error rate

The calculation method is:

{P P}_{b b}^{((22))} = = {P P}_{22} = = Q Q ((\sqrt{22 {γ γ}_{i i}})) - - - - - - ((22))

For QPSK, 16-QAM, and 64-QAM modulation, the bit error rate in Gaussian white noise environment

The calculation method is as follows:

{P P}_{b b}^{(({M m}_{i i}))} \approx \approx \frac{11}{{log log}_{22} {M m}_{i i}} \cdot \cdot {P P}_{s the s,, i i} (({M m}_{i i},, {γ γ}_{i i})) - - - - - - ((33))

Among them, M _i represents the number of types of symbols sent by the specified modulation mode, and the symbol error rate P _{s, i} (M _i , γ _i ):

{P P}_{s the s,, i i} (({M m}_{i i},, {γ γ}_{i i})) = = 11 - - {[[11 - - 22 \cdot &Center Dot; ((11 - - \frac{11}{\sqrt{{M m}_{i i}}})) \cdot &Center Dot; Q Q ((\sqrt{\frac{33}{{M m}_{i i} - - i i} \cdot \cdot {γ γ}_{i i}}))]]}^{22} - - - - - - ((44))

where _γi is the average signal-to-noise ratio per symbol,

E _b,i represents the average energy of a symbol, N _0,i represents the power spectral density of the noise.

Among them, the Q function is

Q (x) = {&Integral;}_{x}^{\infty} \frac{1}{\sqrt{2 π}} &Center Dot; e^{- {the y}^{2} / 2} dy - - - (5)

Then calculate the frame error rate

Assuming that the Viterbi hard-decision decoding method is used, the calculated transmission mode is m, and the integer m represents the serial number corresponding to the eight transmission modes defined in the IEEE 802.11a standard. l ₁ is the total number of preamble symbols and PLCP header bits, and l ₂ is The total number of bits of the MPDU, the frame error rate upper limit formula of the data packet whose length is l=l ₁ +l ₂ bits is as follows:

{P P}_{e e,, i i}^{m m} ((l l)) \leq \leq 11 - - {[[11 - - {P P}_{u u}^{m m} (({l l}_{11}))]]}^{{l l}_{11}} {[[11 - - {P P}_{u u}^{m m} (({l l}_{22}))]]}^{{l l}_{22}} - - - - - - ((66))

{P P}_{u u}^{m m} ((l l)) \leq \leq {Σ Σ}_{d d = = {d d}_{free free}}^{\infty \infty} {a a}_{d d} \cdot \cdot {P P}_{d d} - - - - - - ((77))

In formula (7),

is the joint bound of the first-event error probability,

In formula (8), d _free is the free distance of the convolutional code in mode m; a _d is the total number of error events with weight d; P _d is the error path with distance d from the correct path selected by the Viterbi decoder The probability of , p _m is the bit error rate using the transmission mode m;

Step 2: Calculate the minimum contention window according to the frame error rate P _e,i corresponding to the transmission mode selected by the data packet in step 1

Include the following steps:

\overset{&OverBar; &OverBar;}{{P P}_{e e,, i i}} ((t t + + 11)) = = {β β}_{11} \cdot \cdot \overset{&OverBar; &OverBar;}{{P P}_{e e,, i i}} ((t t)) + + ((11 - - {β β}_{11})) \cdot &Center Dot; {P P}_{e e,, i i} - - - - - - ((99))

In formula (9), β ₁ represents a parameter whose value is between 0 and 1, Indicates the average frame error rate of node i at time t,

Indicates the average frame error rate of node i at time t+1;

{τ τ}^{* *}_{ap ap,, i i} = = \frac{11}{\sqrt{\frac{{T T}_{c c}^{* *}}{22} \cdot &Center Dot; N N}} - - - - - - ((1010))

In formula (10)

Represents the approximate optimal operating point of node i, N represents the number of nodes waiting to send data (N=k),

Indicates the average duration of the normalized conflict;

{T T}_{c c}^{* *} = = \frac{{T T}_{c c}}{σ σ} - - - - - - ((1111))

In the formula, T _c represents the average time for the channel to perceive the collision duration, T _c =T _DIFS +T _RTS , T _DIFS represents the time length of the frame interval DIFS, T _RTS represents the time length for transmitting RTS packets, and σ represents an empty time slot;

{p p}^{* *}_{c c,, i i} \approx \approx 11 - - \frac{{e e}^{- - 11 / / K K}}{((11 - - {τ τ}_{ap ap,, i i}^{* *}))} - - - - - - ((1212))

In formula (12), p ^* _{c, i} represents the collision probability of node i,

K = \sqrt{T_{c}^{*} / 2};

{p p}_{f f,, i i}^{* *} = = {p p}^{* *}_{c c,, i i} + + ((11 - - {p p}^{* *}_{c c,, i i})) \cdot &Center Dot; \overset{&OverBar; &OverBar;}{{P P}_{e e,, i i}} - - - - - - ((1313))

In formula (13)

Indicates the total error probability of node i;

{W W}_{i i}^{* *} \approx \approx \frac{22 ((11 - - 22 {p p}_{f f,, i i}^{* *}))}{{τ τ}^{* *}_{ap ap,, i i} \cdot \cdot ((11 - - 22 {p p}_{f f,, i i}^{* *})) + + {τ τ}^{* *}_{ap ap,, i i} \cdot \cdot {p p}_{f f,, i i}^{* *} \cdot \cdot [[11 - - {((22 {p p}_{f f,, i i}^{* *}))}^{{m m}_{i i}}]]} - - - - - - ((1414))

In formula (14) Indicates the minimum contention window size of node i, m _i indicates the number of back-off time slots for this transmission of node i, and the minimum contention window for the next channel competition after node i successfully sends this packet can be calculated by using formula (14) ;

Step 3: Compare the calculated optimal transmission mode with the minimum contention window Write the newly introduced transmission mode field and the minimum contention window field in the CTS header respectively, and the receiving node sends the CTS packet;

Step 4: If the receiving node receives the data packet correctly, it will return an ACK confirmation packet to the sending node; if the receiving node does not receive the data packet correctly, it will not respond.